[Skip to Navigation]
Sign In
Figure.  Modified Rankin Scale Score at 90 Days Among All Patients and Patients in Clinical Mismatch, Target Perfusion Mismatch, and Undetermined Profile Subgroups
Modified Rankin Scale Score at 90 Days Among All Patients and Patients in Clinical Mismatch, Target Perfusion Mismatch, and Undetermined Profile Subgroups

All patients category includes 504 participants; clinical mismatch subgroup, 295 participants; target perfusion mismatch subgroup, 359 participants; and undetermined profile subgroup, 132 participants. Diagonal blue lines show the shift in favorable and unfavorable outcomes.

Table 1.  Baseline Patient Characteristics
Baseline Patient Characteristics
Table 2.  Baseline Characteristics of Patients With Available vs Undetermined Imaging Profiles
Baseline Characteristics of Patients With Available vs Undetermined Imaging Profiles
Table 3.  Likelihood of Ordinal Change in 90-Day Modified Rankin Scale Scores
Likelihood of Ordinal Change in 90-Day Modified Rankin Scale Scores
Table 4.  Treatment Benefit Based on Analysis of Ordinal Change in 90-Day Modified Rankin Scale Scoresa
Treatment Benefit Based on Analysis of Ordinal Change in 90-Day Modified Rankin Scale Scoresa
1.
Nogueira  RG, Jadhav  AP, Haussen  DC,  et al; DAWN Trial Investigators.  Thrombectomy 6 to 24 hours after stroke with a mismatch between deficit and infarct.   N Engl J Med. 2018;378(1):11-21. doi:10.1056/NEJMoa1706442 PubMedGoogle ScholarCrossref
2.
Albers  GW, Marks  MP, Kemp  S,  et al; DEFUSE 3 Investigators.  Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging.   N Engl J Med. 2018;378(8):708-718. doi:10.1056/NEJMoa1713973 PubMedGoogle ScholarCrossref
3.
Jovin  TG, Chamorro  A, Cobo  E,  et al; REVASCAT Trial Investigators.  Thrombectomy within 8 hours after symptom onset in ischemic stroke.   N Engl J Med. 2015;372(24):2296-2306. doi:10.1056/NEJMoa1503780 PubMedGoogle ScholarCrossref
4.
Goyal  M, Demchuk  AM, Menon  BK,  et al; ESCAPE Trial Investigators.  Randomized assessment of rapid endovascular treatment of ischemic stroke.   N Engl J Med. 2015;372(11):1019-1030. doi:10.1056/NEJMoa1414905 PubMedGoogle ScholarCrossref
5.
Martins  SO, Mont’Alverne  F, Rebello  LC,  et al; RESILIENT Investigators.  Thrombectomy for stroke in the public health care system of Brazil.   N Engl J Med. 2020;382(24):2316-2326. doi:10.1056/NEJMoa2000120 PubMedGoogle ScholarCrossref
6.
Mocco  J, Siddiqui  AH, Fiorella  D,  et al.  POSITIVE: perfusion imaging selection of ischemic stroke patients for endovascular therapy.   J Neurointerv Surg. Published online February 25, 2021. doi:10.1136/neurintsurg-2021-017315 PubMedGoogle Scholar
7.
Jovin  T. Thrombectomy for anterior circulation stroke beyond 6 hours from time last known well: the AURORA (Analysis of Pooled Data From Randomized Studies of Thrombectomy More Than 6 Hours After Last Known Well) Collaboration. Abstract presented at: 4th European Stroke Organisation Conference; May 16, 2018; Gothenburg, Sweden.
8.
Campbell  BCV, Mitchell  PJ, Kleinig  TJ,  et al; EXTEND-IA Investigators.  Endovascular therapy for ischemic stroke with perfusion-imaging selection.   N Engl J Med. 2015;372(11):1009-1018. doi:10.1056/NEJMoa1414792 PubMedGoogle ScholarCrossref
9.
Saver  JL, Goyal  M, Bonafe  A,  et al; SWIFT PRIME Investigators.  Stent-retriever thrombectomy after intravenous t-PA vs. t-PA alone in stroke.   N Engl J Med. 2015;372(24):2285-2295. doi:10.1056/NEJMoa1415061 PubMedGoogle ScholarCrossref
10.
Olivot  JM, Albucher  JF, Guenego  A,  et al; FRAME Investigators.  Mismatch profile influences outcome after mechanical thrombectomy.   Stroke. 2021;52(1):232-240. doi:10.1161/STROKEAHA.120.031929 PubMedGoogle ScholarCrossref
11.
Campbell  BCV, Ma  H, Ringleb  PA,  et al; EXTEND, ECASS-4, and EPITHET Investigators.  Extending thrombolysis to 4.5-9 h and wake-up stroke using perfusion imaging: a systematic review and meta-analysis of individual patient data.   Lancet. 2019;394(10193):139-147. doi:10.1016/S0140-6736(19)31053-0 PubMedGoogle ScholarCrossref
12.
Leslie-Mazwi  TM, Hamilton  S, Mlynash  M,  et al.  DEFUSE 3 non-DAWN patients.   Stroke. 2019;50(3):618-625. doi:10.1161/STROKEAHA.118.023310 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    July 26, 2021

    Assessment of Optimal Patient Selection for Endovascular Thrombectomy Beyond 6 Hours After Symptom Onset: A Pooled Analysis of the AURORA Database

    Author Affiliations
    • 1Department of Neurology and Neurological Sciences, Stanford University, Stanford, California
    • 2Stanford Stroke Center, Stanford University, Stanford, California
    • 3Bright Research Partners, Minneapolis, Minnesota
    • 4Department of Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona
    • 5Department of Neurology, Marcus Stroke and Neuroscience Center, Grady Memorial Hospital, Emory University School of Medicine, Atlanta, Georgia
    • 6Federal University of Rio Grande do Sul, Porto Alegre, Brazil
    • 7Department of Clinical Neurosciences, Calgary Stroke Program, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
    • 8Department of Radiology, Calgary Stroke Program, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada
    • 9Stroke Unit, Department of Neurology, Hospital Vall d’Hebron, Barcelona, Spain
    • 10Department of Neurosurgery, Prisma Health–Upstate, Greenville, South Carolina
    • 11Department of Neurology, University of California, Los Angeles, Los Angeles
    • 12UCLA Stroke Center, University of California, Los Angeles, Los Angeles
    • 13Department of Radiology and Stanford Stroke Center, Stanford University, Stanford, California
    • 14Department of Neurology, Cooper University Health Care, Camden, New Jersey
    JAMA Neurol. 2021;78(9):1064-1071. doi:10.1001/jamaneurol.2021.2319
    Key Points

    Question  What is the optimal imaging approach for identification of patients with ischemic stroke who may benefit from endovascular thrombectomy beyond 6 hours after they were last known well?

    Findings  In this pooled analysis of 6 randomized clinical trials including 505 patients with ischemic stroke, the receipt of endovascular thrombectomy was associated with a reduction in disability among patients with an imaging profile mismatch between clinical defect vs size of early infarction or size of perfusion lesion vs size of early infarction; however, this reduction in disability was not observed in the smaller group of patients with an undetermined imaging profile. Greater benefit was observed among patients with one of the mismatch profiles compared with those with an undetermined imaging profile.

    Meaning  In this study, within a 6- to 24-hour treatment interval, the performance of endovascular thrombectomy among patients who had either mismatch imaging profile was associated with improvements in clinical outcomes; additional studies are needed to clarify the treatment benefit among patients with other imaging profiles.

    Abstract

    Importance  The optimal imaging approach for identifying patients who may benefit from endovascular thrombectomy (EVT) beyond 6 hours after they were last known well is unclear. Six randomized clinical trials (RCTs) have evaluated the efficacy of EVT vs standard medical care among patients with ischemic stroke.

    Objective  To assess the benefits of EVT among patients with 3 baseline imaging profiles using a pooled analysis of RCTs.

    Data Sources  The AURORA (Analysis of Pooled Data from Randomized Studies of Thrombectomy More Than 6 Hours After Last Known Well) Collaboration pooled patient-level data from the included clinical trials.

    Study Selection  An online database search identified RCTs of endovascular stroke therapy published between January 1, 2010, and March 1, 2021, that recruited patients with ischemic stroke who were randomized between 6 and 24 hours after they were last known well.

    Data Extraction/Synthesis  Data from the final locked database of each study were provided. Data were pooled, and analyses were performed using mixed-effects modeling with fixed effects for parameters of interest.

    Main Outcomes and Measures  The primary outcome was reduction in disability measured by the modified Rankin Scale at 90 days. An evaluation was also performed to examine whether the therapeutic response differed based on imaging profile among patients who received treatment based on the time they were last known well. Treatment benefits were assessed among a clinical mismatch subgroup, a target perfusion mismatch subgroup, and an undetermined profile subgroup. The primary end point was assessed among these subgroups and during 3 treatment intervals (tercile 1, 360-574 minutes [6.0-9.5 hours]; tercile 2, 575-762 minutes [9.6-12.7 hours]; and tercile 3, 763-1440 minutes [12.8-24.0 hours]).

    Results  Among 505 eligible patients, 266 (mean [SD] age, 68.4 [13.8] years; 146 women [54.9%]) were assigned to the EVT group and 239 (mean [SD] age, 68.7 [13.7] years; 126 men [52.7%]) were assigned to the control group. Among 295 patients in the clinical mismatch subgroup and 359 patients in the target perfusion mismatch subgroup, EVT was associated with reductions in disability at 90 days vs no EVT (clinical mismatch subgroup, odds ratio [OR], 3.57; 95% CI, 2.29-5.57; P < .001; target perfusion mismatch subgroup, OR, 3.13; 95% CI, 2.10-4.66; P = .001). Statistically significant benefits were observed in all 3 terciles for both subgroups, with the highest OR observed for tercile 3 (clinical mismatch subgroup, OR, 4.95; 95% CI, 2.20-11.16; P < .001; target perfusion mismatch subgroup, OR, 5.01; 95% CI, 2.37-10.60; P < .001). A total of 132 patients (26.1%) had an undetermined imaging profile and no significant treatment benefit (OR, 1.59; 95% CI, 0.82-3.06; P = .17). The interaction between treatment effects for the clinical and target perfusion mismatch subgroups vs the undetermined profile subgroup was significant (OR, 2.28; 95% CI, 1.11-4.70; P = .03).

    Conclusions and Relevance  In this study, EVT was associated with similar benefit among patients in the clinical mismatch and target perfusion mismatch subgroups during the 6- to 24-hour treatment interval. These findings support EVT as a treatment for patients meeting the criteria for either of the imaging mismatch profiles within the 6- to 24-hour interval.

    Introduction

    The publications of the DAWN (Diffusion Weighted Imaging or Computerized Tomography Perfusion Assessment With Clinical Mismatch in the Triage of Wake Up and Late Presenting Strokes Undergoing Neurointervention)1 and DEFUSE 3 (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3)2 clinical trials in 2018 led to new guidelines that recommend endovascular thrombectomy (EVT) for eligible patients with an anterior circulation large artery occlusion within a 6- to 24-hour interval. The DAWN study1 selected patients who could be treated from 6 to 24 hours based on a clinical mismatch profile that included criteria for age, ischemic core size based on computed tomographic perfusion (CTP) or diffusion-weighted magnetic resonance imaging (MRI) scans, and baseline National Institutes of Health Stroke Scale (NIHSS) score.1 The DEFUSE 3 study2 used a shorter treatment interval (6-16 hours) and selected patients with a target perfusion mismatch profile using MRI or CTP scans.2 The REVASCAT (Randomized Trial of Revascularization With Solitaire FR Device Versus Best Medical Therapy in the Treatment of Acute Stroke Due to Anterior Circulation Large Vessel Occlusion Presenting Within 8 Hours of Symptom Onset),3 ESCAPE (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion With Emphasis on Minimizing CT to Recanalization Times),4 and RESILIENT (Randomization of Endovascular Treatment With Stent Retriever and/or Thromboaspiration vs Best Medical Therapy in Acute Ischemic Stroke Due to Large Vessel Occlusion Trial)5 clinical trials had shorter maximum treatment intervals (8, 12, and 8 hours, respectively). These 3 studies did not require perfusion imaging, but CTP scans were obtained for a subset of patients (eg, the REVASCAT study3 required CTP scans for patients who were enrolled after 4.5 hours, but these scans were mainly used as an aid to Alberta Stroke Programme Early CT Score [ASPECTS] interpretation because the entry criteria were based on the CTP ASPECTS results rather than CTP lesion volumes). The POSITIVE (Perfusion Imaging Selection of Ischemic Stroke Patients for Endovascular Therapy)6 clinical trial required perfusion imaging and used mismatch criteria based on a visual assessment of salvageable tissue and a treatment interval of 0 to 12 hours.

    Our aims were to assess the data from all 6 clinical trials1-6 among patients enrolled beyond 6 hours after they were last known well (defined as the time at which the patient was last known to be without the symptoms of the current stroke) using the same perfusion imaging software program and to evaluate the treatment benefit among 3 subgroups: (1) patients who met the clinical imaging mismatch profile used in the DAWN study1 (clinical mismatch subgroup), (2) patients who met the target perfusion imaging mismatch profile used in the DEFUSE 3 study2 (target perfusion mismatch subgroup), and (3) patients with an undetermined imaging mismatch profile (undetermined profile subgroup; ie, patients for whom the imaging profiles could not be determined because a technically adequate CT or MRI perfusion study was not performed). We assessed the primary and secondary end points among these 3 patient subgroups and during 3 treatment intervals (tercile 1 was 360-574 minutes [6.0-9.5 hours], tercile 2 was 575-762 minutes [9.6-12.7 hours], and tercile 3 was 763-1440 minutes [12.8-24.0 hours]).

    Methods

    The AURORA (Analysis of Pooled Data from Randomized Studies of Thrombectomy More Than 6 Hours After Last Known Well) study is a pooled analysis of all 6 published randomized clinical trials (RCTs)1-6 that allowed inclusion of patients with anterior circulation ischemic stroke who were randomized beyond 6 hours after they were last known well and were treated with second-generation neurothrombectomy devices.7 To identify the included studies, we searched online databases, including MEDLINE and PubMed, for RCTs of endovascular stroke therapy published between January 1, 2010, and March 1, 2021. This study followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.

    Of the 505 patients in the AURORA database (266 assigned to the EVT group and 239 assigned to the control group), 1 patient did not have a 90-day modified Rankin Scale (mRS) score available; therefore, 504 patients were included in the assessment of 90-day outcomes. A total of 372 patients had all of the following data available: baseline time to maximum of greater than 6 seconds perfusion and ischemic core lesion volumes, age, baseline NIHSS score (range, 0-42, with higher scores indicating greater stroke severity), and 90-day mRS score (range, 0-5, with 0 indicating no symptoms and 5 indicating severe disability). Therefore, both imaging profiles (clinical and target perfusion mismatch) could be assessed among this group. Among those 372 patients, 180 participants were enrolled in DEFUSE 3,2 163 in DAWN,1 12 in RESILIENT,5 10 in ESCAPE,4 4 in REVASCAT,3 and 3 in POSITIVE.6 Among 132 patients with an undetermined mismatch profile, 43 participants were enrolled in DAWN,1 38 in ESCAPE,4 18 in POSITIVE,6 17 in REVASCAT,3 14 in RESILIENT,5 and 2 in DEFUSE 3.2

    We divided the AURORA patients into subgroups based on the 3 imaging profiles after processing data from CTP and diffusion-weighted MRI and/or perfusion scans. Rapid Processing of Perfusion and Diffusion (RAPID) software, version 4.6 (iSchemaView) was used to process data from patients for whom RAPID processing was not performed during the original clinical trial (ie, patients enrolled in the ESCAPE,4 POSITIVE,6 and REVASCAT3 studies). The original RAPID volumes obtained at the time of enrollment were used from the other clinical trials.1,2,5 As in the DEFUSE 32 and DAWN1 clinical trials, the ischemic core volume was estimated from the prerandomization scan as a relative cerebral blood flow of less than 30% of normal brain blood flow or an apparent diffusion coefficient of less than 620 μm2/s as measured by MRI scan. Critically hypoperfused tissue was estimated as the volume of tissue with a time to maximum of greater than 6 seconds lesion volume. The mismatch volume (ie, the estimated penumbral volume) was calculated as the critically hypoperfused tissue volume minus the ischemic core volume.

    Patients’ ASPECTS results (range, 0-10, with 10 indicating a healthy brain and 1 point subtracted for any evidence of early ischemic change in each of the defined brain regions) were assessed by the Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials (HERMES) core laboratory for the REVASCAT3 and ESCAPE4 clinical trials and by individual study core laboratories for all other clinical trials.1,2,5,6 Patients with clinical mismatch profiles (ie, a mismatch between clinical defect and size of early infarction) were defined as those 80 years or older with an NIHSS score of 10 or higher and an ischemic core volume of less than 21 mL, those younger than 80 years with an NIHSS score of 10 or higher and an ischemic core volume of less than 31 mL, and those younger than 80 years with an NIHSS score of 20 or higher and an ischemic core volume of 31 mL to less than 51 mL. Patients with target perfusion mismatch profiles (ie, a mismatch between size of perfusion lesion and size of early infarction) were defined as those with an ischemic core volume of less than 70 mL, a ratio of volume of critically hypoperfused tissue (time to maximum >6 seconds lesion volume) to ischemic core volume of 1.8 or higher, and an absolute volume of potentially reversible ischemia (estimated penumbra) of 15 mL or higher. A flowchart of the 3 imaging subgroups is available in the eFigure in Supplement 1.

    The primary outcome was the ordinal change in 90-day mRS scores, and the secondary outcome was the rate of mRS scores of 0 to 2 at 90 days. Both analyses used an ordinal logistic regression model. We adjusted for age, sex, baseline stroke severity (based on NIHSS score), site of occlusion (internal carotid artery vs M1 segment of the middle cerebral artery vs M2 segment of the middle cerebral artery, ASPECTS result, and time from stroke onset to randomization). To account for between-study variance, we used mixed-effects modeling with random effects for parameters of interest. We then used mixed-effects ordinal logistic regression models to calculate common odds ratios (ORs) for the primary and secondary outcomes in the entire population and in the 3 prespecified subgroups based on imaging profiles. We also examined the interaction between treatment effects among patients with undetermined imaging profiles vs patients for whom both imaging profiles (clinical and target perfusion mismatch) were available.

    Statistical analyses were performed using SAS software, version 9.4 (SAS Institute), and R software, version 3.5.2 (R Foundation for Statistical Computing). Statistical significance was set at P < .05.

    Results

    Among 505 eligible patients, 266 individuals (mean [SD] age, 68.4 [13.8] years; 146 women [54.9%] and 120 men [45.1%]) were assigned to the EVT group, and 239 individuals (mean [SD] age, 68.7 [13.7] years; 113 women [47.3%] and 126 men [52.7%]) were assigned to the control group (Table 1). Overall, 378 of 503 patients (75.1%) had hypertension, and 247 of 494 patients (50.0%) had hyperlipidemia. The mean (SD) baseline NIHSS score was 16.4 (5.3) in the EVT group and 17.2 (5.9) in the control group, and the mean (SD) baseline ASPECTS result was 7.8 (1.3) in the EVT group and 7.5 (1.6) in the control group.

    A total of 373 patients (73.9%) had imaging profiles (clinical mismatch and/or perfusion mismatch) that could both be assessed, and 132 patients (26.1%) had undetermined imaging profiles (Table 2). Among those with available vs undetermined imaging profiles, baseline NIHSS scores (mean [SD], 16.8 [5.8] vs 16.7 [4.8], respectively) and ASPECTS results (mean [SD], 7.6 [1.4] vs 7.7 [1.5]) were similar.

    Of 372 patients for whom both imaging profiles and 90-day mRS scores could be assessed, 359 patients (96.5%) met the criteria for the target perfusion mismatch profile, 295 patients (79.3%) met the criteria for the clinical mismatch profile, and 283 patients (76.1%) met the criteria for both the target perfusion and clinical mismatch profiles. A total of 64 patients met the target perfusion mismatch criteria but not the clinical mismatch criteria, and 12 patients met the clinical mismatch criteria but not the target perfusion mismatch criteria. Only 1 patient with adequate perfusion imaging did not meet the criteria for clinical or target perfusion mismatch profiles; this patient was therefore included in the undetermined profile subgroup.

    The results of the ordinal logistic regression model for treatment effect are shown in Table 3, and the distribution of mRS scores at 90 days are shown in the Figure. Both the clinical and target perfusion mismatch subgroups experienced a treatment benefit in terms of lower overall disability and a higher likelihood of functional independence at 90 days (clinical mismatch subgroup: OR, 3.57 [95% CI, 2.29-5.57]; P < .001; target perfusion mismatch subgroup: OR, 3.13 [95% CI, 2.10-4.66]; P = .001). The results for the undetermined profile subgroup were inconclusive, revealing a nonsignificant favorable numerical difference in overall disability and functional independence at 90 days (OR, 1.59; 95% CI, 0.82-3.06; P = .17). The interaction between treatment effect among patients with an undetermined imaging profile vs those with either a clinical or target perfusion mismatch profile was significant (OR for interaction effect, 2.28; 95% CI, 1.11-4.70; P = .03), with a greater treatment benefit found among those with a mismatch profile. The results of the ordinal logistic regression model for the secondary end point of an mRS score of 0 to 2 at 90 days are shown in the eTable in Supplement 1.

    Among those in the undetermined profile subgroup, 128 patients had an NIHSS score of 10 or higher. The OR for the primary outcome among these patients was 1.52 (95% CI, 0.78-2.98; P = .22). Among 38 study participants with a baseline NIHSS score of 9 or lower, the OR was 1.81 (95% CI, 0.47-7.02; P = .38). The results of the secondary analysis (mRS score of 0-2 at 90 days) were similar to those of the primary analysis and are shown in the eTable in Supplement 1. The results of the primary analysis based on the time the patient was last known well are shown in Table 4. Statistically significant benefits were observed in all 3 time terciles for both the clinical mismatch and target perfusion mismatch subgroups, with the highest OR observed for time tercile 3 (clinical mismatch subgroup: OR, 4.95; 95% CI, 2.20-11.16; P < .001; target perfusion mismatch subgroup: OR, 5.01; 95% CI, 2.37-10.60; P < .001). No statistically significant differences in mortality were found between the EVT and control groups among any of the 3 subgroups or the full population (eg, clinical mismatch subgroup: OR, 0.94 [95% CI, 0.48-1.84]; target perfusion mismatch subgroup: OR, 0.83 [95% CI, 0.45-1.53]; undetermined profile subgroup: OR, 1.34 [95% CI, 0.45-3.96]).

    For the cohort of 411 participants with an NIHSS score of 6 or higher and an ASPECTS result of 6 or higher, the OR was 2.45 (95% CI, 1.70-3.54; P < .001). Among 426 participants with an NIHSS score of 6 or higher and an ischemic core volume of up to 70 mL, the OR was 2.98 (95% CI, 2.07-4.29; P < .001). Among patients in the undetermined profile subgroup with an NIHSS score of 6 or higher and an ASPECTS result of 6 or higher, the OR was 1.72 (95% CI, 0.82-3.58; P = .15). Overall, only 42 patients had an ASPECTS result lower than 6; among those with either a clinical or perfusion mismatch profile, 31 patients had an ASPECTS result lower than 6. Among all 505 patients in the AURORA database, 124 patients (24.6%) were selected using MRI data, and the remaining patients were selected using CT or CTP data.

    Discussion

    The data from this pooled analysis suggest that receipt of EVT during an extended period after symptom onset is associated with similar benefit among patients with a clinical mismatch profile and those with a target perfusion mismatch profile throughout the 6- to 24-hour treatment interval. In the DEFUSE 3 clinical trial,2 which used the target perfusion mismatch profile for selection, patients did not receive treatment beyond 16 hours. However, in the DAWN clinical trial,1 which used the clinical mismatch criteria, patients received treatment for up to 24 hours. To our knowledge, this study provides the first data on the benefits of endovascular therapy among patients with a target perfusion mismatch profile beyond 16 hours after they were last known well. The associated benefits for these patients were substantial (Table 4).

    In general, target perfusion mismatch criteria identify a larger number of patients than clinical mismatch criteria (eg, 359 patients vs 295 patients, respectively, in the AURORA database). Patients who qualify as having target perfusion mismatch and nonclinical mismatch profiles include those with larger ischemic core volumes (>30 mL for those aged ≥80 years and >50 mL for those aged <80 years) and those presenting with a baseline NIHSS score of 6 to 9.

    A target perfusion mismatch approach for patient selection in endovascular RCTs was first used in the EXTEND-IA (Extending the Time for Thrombolysis in Emergency Neurological Deficits–Intra-arterial)8 and SWIFT PRIME (Solitaire With the Intention for Thrombectomy as Primary Endovascular Treatment)9 studies, both of which enrolled patients within 6 hours of symptom onset. The recent FRAME (French Acute Cerebral Multimodal Imaging to Select Patient for Mechanical Thrombectomy) study10 also assessed perfusion imaging profiles early in the treatment interval and reported that perfusion profiles were associated with clinical response to EVT. A perfusion mismatch approach was used in the EXTEND clinical trial,11 which reported a treatment benefit associated with intravenous alteplase therapy vs placebo among patients enrolled within 4.5 to 9.0 hours of symptom onset. A pooled analysis of the EXTEND study and 2 additional clinical trials that obtained perfusion imaging before randomizing patients to receive intravenous alteplase therapy vs placebo beyond 4.5 hours after stroke onset reported that the rate of excellent functional outcomes was higher in the alteplase group than the placebo group (36% vs 26%, respectively; adjusted OR, 2.06; 95% CI, 1.17-3.62; P = .01).11 Among patients without mismatch imaging profiles, no significant differences were observed.

    Using a 24-hour interval for endovascular therapy and treatment of patients who meet either the clinical or target perfusion mismatch profile will increase the number of patients who can be treated. In the present study, patients who had an undetermined mismatch profile because perfusion imaging was not performed did not experience a treatment benefit, and a significant interaction between treatment effect and mismatch status was found. The AURORA Collaboration was not able to assess treatment benefits in nonmismatched patients because these patients were excluded from the 2 largest studies1,2 in the AURORA database, and only 1 nonmismatched patient was included in the other studies.3-6 Studies of patients with larger perfusion-based ischemic core lesions, such as the SELECT 2 (A Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; ClinicalTrials.gov identifier: NCT03876457) clinical trial, and studies that use nonperfusion-based imaging, such as the MR CLEAN-LATE (Multicenter Randomized Clinical Trial of Endovascular Treatment of Acute Ischemic Stroke in The Netherlands for Late Arrivals; isrctn.org identifier: ISRCTN19922220) and RESILIENT-Extend (Randomization of Endovascular Treatment With Stent Retriever and/or Thromboaspiration vs Best Medical Therapy in Acute Ischemic Stroke Due to Large Vessel Occlusion Trial in the Extended Time Window; ClinicalTrials.gov identifier: NCT04256096) clinical trials, are currently ongoing.

    Limitations

    This study has limitations. It is possible that the treatment benefit may be greater in those patients who qualified as having target perfusion mismatch, but not clinical mismatch, because the ischemic core volume was too large to qualify for clinical mismatch (>30 mL for those aged ≥80 years and >50 mL for those aged <80 years) vs those presenting with baseline NIHSS scores of 6 to 9. This issue was highlighted in a DEFUSE 3 substudy12 that reported a significant treatment benefit among patients who qualified as having target perfusion mismatch, but not clinical mismatch, based on ischemic core size but found no significant benefit among patients who did not qualify for clinical mismatch based on low baseline NIHSS scores. The analysis of 132 patients with an unknown mismatch profile is underpowered, and the favorable numerical differences that were observed in functional independence at 90 days could be clinically relevant if the analysis were adequately powered. In addition, the study does not include patients who received treatment less than 6 hours after stroke onset.

    Conclusions

    The findings of this pooled analysis support EVT for the treatment of patients who meet criteria for either the clinical or target perfusion imaging mismatch profile throughout the 6- to 24-hour treatment interval.

    Back to top
    Article Information

    Accepted for Publication: April 30, 2021.

    Published Online: July 26, 2021. doi:10.1001/jamaneurol.2021.2319

    Correction: This article was corrected on September 9, 2021, to correct an error in the Figure. In the “All patients” section of the Figure, the left-hand connector line should have extended to the left side of the light blue box, such that modified Rankin Scale scores 3 and 4 are included between the 2 connector lines in the control group. The Figure has been corrected accordingly.

    Corresponding Author: Gregory W. Albers, MD, Department of Neurology and Neurological Sciences, Stanford University, 453 Quarry Rd, Stanford, CA 94305 (albers@stanford.edu).

    Author Contributions: Drs Albers and Brown had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Albers, Lansberg, Turk, Liebeskind, Jovin, Nogueira.

    Acquisition, analysis, or interpretation of data: Albers, Lansberg, Brown, Jadhav, Haussen, Martins, Rebello, Demchuk, Goyal, Ribo, Liebeskind, Heit, Marks, Nogueira.

    Drafting of the manuscript: Albers.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Brown.

    Obtained funding: Demchuk, Jovin, Nogueira.

    Administrative, technical, or material support: Lansberg, Jadhav, Demchuk, Turk, Liebeskind, Heit, Nogueira.

    Supervision: Nogueira.

    Conflict of Interest Disclosures: Dr Albers reported receiving consulting fees from Genentech and owning equity in iSchemaView outside the submitted work. Dr Lansberg reported receiving grants from the National Institutes of Health during the conduct of the study. Dr Brown reported receiving consulting fees from the University of Pittsburgh during the conduct of the study and consulting fees from Medtronic outside the submitted work. Dr Haussen reported receiving consulting fees from Cerenovus, Stryker Neurovascular, and Vesalio during the conduct of the study and owning stock in Viz.ai outside the submitted work. Dr Martins reported receiving consulting and speaking fees from Medtronic and speaking fees from Penumbra outside the submitted work. Dr Demchuk reported receiving consulting fees and honoraria from Medtronic and honoraria from Boehringer Ingelheim during the conduct of the study and owning stock in Circle NVI and a patent for stroke imaging software (licensed to Circle NVI) outside the submitted work. Dr Goyal reported receiving consulting fees from Medtronic, Mentice, MicroVention, and Stryker Neurovascular and owning patents for acute stroke diagnostic systems (licensed to GE Healthcare) and intracranial access systems (licensed to MicroVention) outside the submitted work. Dr Ribo reported receiving consulting fees from aptaTargets, Medtronic, Neuravi/Cerenovus, Stryker Neurovascular, and Vesalio and owning stock in Anaconda Biomed SL and methinks Technologies outside the submitted work. Dr Turk reported receiving grants from Canon Medical Systems, Medtronic, Penumbra, and Stryker Neurovascular during the conduct of the study; owning stock in Bendit Technologies, Cerebrotech Medical Systems, EndoStream Medical, Imperative Care, Instylla, Q’Apel Medical, Spinnaker, Synchron, Truvic Medical, and Vastrax; receiving consulting fees from Medtronic, MicroVention, Penumbra, and Stryker Neurovascular; and serving as the chief medical officer of Corindus Vascular Robotics and Imperative Care outside the submitted work. Dr Liebeskind reported receiving consulting fees from Cerenovus, Genentech, Medtronic, Rapid Medical, and Stryker Neurovascular during the conduct of the study. Dr Heit reported receiving consulting or advisory fees from iSchemaView, Medtronic, and MicroVention and owning equity in ThrombX Medical outside the submitted work. Dr Marks reported receiving grants from the National Institute of Neurological Disorders and Stroke during the conduct of the study; receiving personal fees from Bayer, Medtronic, Neuravi/Cerenovus, and Servier Laboratories; serving on the steering committees of the SWIFT, PRIME, and ARISE clinical trials; and owning a patent for a thrombectomy device (licensed to ThrombX Medical) and stock in ThrombX Medical outside the submitted work. Dr Jovin reported receiving grants from Stryker Neurovascular during the conduct of the study; receiving grants from Medtronic; receiving consulting/advisory fees from Anaconda Biomed SL, Blockade Medical, Cerenovus, Contego Medical, Corindus Vascular Robotics, FreeOx Biotech, methinks Technologies, Route 92 Medical, and Viz.ai; and owning stock in Anaconda Biomed SL, Blockade Medical, Corindus Vascular Robotics, FreeOx Biotech, Medtronic, Neuravi/Cerenovus, Rapid Medical, Route 92 Medical, and Viz.ai outside the submitted work. Dr Nogueira reported receiving research funding from Allm, Anaconda Biomed SL, Biogen, Brainomix, Ceretrieve, Corindus Vascular Robotics, Genentech, Medtronic, Neuravi/Cerenovus, Penumbra, phenox, Prolong Pharmaceuticals, Sensome, Stryker Neurovascular, Vesalio, and Viz.ai; receiving consulting fees from Anaconda Biomed SL, Biogen, Cerenovus, Genentech, Imperative Care, Medtronic, phenox, Prolong Pharmaceuticals, and Stryker Neurovascular; and owning stock in Astrocyte Pharmaceuticals, Brainomix, Cerebrotech Medical Systems, Ceretrieve, Corindus Vascular Robotics, Perfuze, Vesalio, and Viz.ai outside the submitted work. No other disclosures were reported.

    Funding/Support: The AURORA collaboration was funded by a grant from Stryker Neurovascular (AURORA Investigators).

    Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Group Information: The AURORA Investigators are listed in Supplement 2.

    References
    1.
    Nogueira  RG, Jadhav  AP, Haussen  DC,  et al; DAWN Trial Investigators.  Thrombectomy 6 to 24 hours after stroke with a mismatch between deficit and infarct.   N Engl J Med. 2018;378(1):11-21. doi:10.1056/NEJMoa1706442 PubMedGoogle ScholarCrossref
    2.
    Albers  GW, Marks  MP, Kemp  S,  et al; DEFUSE 3 Investigators.  Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging.   N Engl J Med. 2018;378(8):708-718. doi:10.1056/NEJMoa1713973 PubMedGoogle ScholarCrossref
    3.
    Jovin  TG, Chamorro  A, Cobo  E,  et al; REVASCAT Trial Investigators.  Thrombectomy within 8 hours after symptom onset in ischemic stroke.   N Engl J Med. 2015;372(24):2296-2306. doi:10.1056/NEJMoa1503780 PubMedGoogle ScholarCrossref
    4.
    Goyal  M, Demchuk  AM, Menon  BK,  et al; ESCAPE Trial Investigators.  Randomized assessment of rapid endovascular treatment of ischemic stroke.   N Engl J Med. 2015;372(11):1019-1030. doi:10.1056/NEJMoa1414905 PubMedGoogle ScholarCrossref
    5.
    Martins  SO, Mont’Alverne  F, Rebello  LC,  et al; RESILIENT Investigators.  Thrombectomy for stroke in the public health care system of Brazil.   N Engl J Med. 2020;382(24):2316-2326. doi:10.1056/NEJMoa2000120 PubMedGoogle ScholarCrossref
    6.
    Mocco  J, Siddiqui  AH, Fiorella  D,  et al.  POSITIVE: perfusion imaging selection of ischemic stroke patients for endovascular therapy.   J Neurointerv Surg. Published online February 25, 2021. doi:10.1136/neurintsurg-2021-017315 PubMedGoogle Scholar
    7.
    Jovin  T. Thrombectomy for anterior circulation stroke beyond 6 hours from time last known well: the AURORA (Analysis of Pooled Data From Randomized Studies of Thrombectomy More Than 6 Hours After Last Known Well) Collaboration. Abstract presented at: 4th European Stroke Organisation Conference; May 16, 2018; Gothenburg, Sweden.
    8.
    Campbell  BCV, Mitchell  PJ, Kleinig  TJ,  et al; EXTEND-IA Investigators.  Endovascular therapy for ischemic stroke with perfusion-imaging selection.   N Engl J Med. 2015;372(11):1009-1018. doi:10.1056/NEJMoa1414792 PubMedGoogle ScholarCrossref
    9.
    Saver  JL, Goyal  M, Bonafe  A,  et al; SWIFT PRIME Investigators.  Stent-retriever thrombectomy after intravenous t-PA vs. t-PA alone in stroke.   N Engl J Med. 2015;372(24):2285-2295. doi:10.1056/NEJMoa1415061 PubMedGoogle ScholarCrossref
    10.
    Olivot  JM, Albucher  JF, Guenego  A,  et al; FRAME Investigators.  Mismatch profile influences outcome after mechanical thrombectomy.   Stroke. 2021;52(1):232-240. doi:10.1161/STROKEAHA.120.031929 PubMedGoogle ScholarCrossref
    11.
    Campbell  BCV, Ma  H, Ringleb  PA,  et al; EXTEND, ECASS-4, and EPITHET Investigators.  Extending thrombolysis to 4.5-9 h and wake-up stroke using perfusion imaging: a systematic review and meta-analysis of individual patient data.   Lancet. 2019;394(10193):139-147. doi:10.1016/S0140-6736(19)31053-0 PubMedGoogle ScholarCrossref
    12.
    Leslie-Mazwi  TM, Hamilton  S, Mlynash  M,  et al.  DEFUSE 3 non-DAWN patients.   Stroke. 2019;50(3):618-625. doi:10.1161/STROKEAHA.118.023310 PubMedGoogle ScholarCrossref
    ×